Probe data showing the presence of waves in Venus's middle and upper atmosphere [Seiff et al., 1989; Seiff and Kirk, 1982> are critically reevaluated and extended to 138 km, near the level of in situ data taken by the Pioneer Venus orbiter. Uncertainties in temperature are determined. They are typically about 0.1 times amplitude, thus supporting the reality of large amplitude oscillations approaching 40 K at 120 km. Growth rates above 100 km follow approximately the inverse square root of density until ''saturation'' occurs (in the sense that lapse rates becomes adiabatic in the expanding segment of the wave). The waves then break at the 120 km level, providing a source for the ''friction'' required in models to match the observed day-night temperature contrast in Venus's lower thermosphere [Seiff, 1982; Bougher, 1984>. The data correlate to an unexpected degree with temperatures from the Pioneer Venus orbiter atmospheric drag (OAD) experiment taken at altitudes of 140 to 165 km, which, especially for the night probe, extend not only the mean temperature structure, but also the oscillation structure of the probe data at the same local Venus time. OAD temperatures depend on local Venus time and altitude, but, in the limited number of observations, appear independent of observing date over periods of up to 11 days, and correlate as described with probe data taken 65 to 137 days earlier. These observations lead to the suggestion that the thermospheric waves are solar-fixed, induced either by the major subsidence across the terminators or as continuations upward of waves in the middle atmosphere. The wave structure in the large probe sounding below 100 km is similar to, but does not quantitatively support the solar-tidal model of Pechmann and Ingersoll, which gives much larger amplitudes and different wave phases. ¿ American Geophysical Union 1991